engleski

Comparison of Genetic and Bees Algorithm in the Finite Element Model Update

This paper investigates and compares genetic algorithms (GA) and bees algorithms (BA) applied to the finite element (FE) model update. The bees algorithm is a new population based optimisation algorithm that mimics the natural foraging of a swarm of bees. The investigation was carried out experimentally on a free-free beam. Five material and geometrical beam parameters were used in this FE model updating procedure. Two objective functions were used: the first one is the residual force vector, while the second consists of experimental and numerical modal parameters only. In order to implement the residual force vector it is necessary to apply model reduction. The effect of model reduction on the process of model updating is investigated. At the end, the values of modal parameters obtained by experimental modal testing were compared with numerical ones obtained by using these two objective functions with GA and BA.

genetic algorithm; bees algorithm; model reduction; finite element model update

nije evidentirano